This is the first resubmission of R01 AG033042. This project focuses on further development of Pittsburgh Compound-B (or PiB) as a positron emission tomography (PET) amyloid imaging agent. Our previous PiB PET efforts resulted in the development of valid simple in vivo methods that provided a solid basis for the use of PiB PET at centers world-wide (including participating ADNI sites). The revised application now includes region-matched comparisons of in vivo PiB retention and post-mortem correlates of amyloid deposition but no longer includes MR image acquisition (diffusion tensor imaging was removed). Amyloid deposition can begin well before the earliest clinical symptoms of Alzheimer's disease (AD). PiB PET has indicated that amyloid deposition begins earliest in frontal and posterior cingulate/precuneus areas of brain, in as many as 25-30% of cognitively normal elders. The primary objective of the proposed R01 is to use PiB PET imaging to establish in vivo thresholds that can be used to distinguish between cognitively normal subjects who have amyloid plaque deposition (PiB+) and those that do not (PiB). The R01 aims will be addressed over 5 years in tandem with an ongoing Merit award PiB PET normal aging study (R37 AG025516 Amyloid Pathology and Cognition in Normal Aging , PI: Klunk). The aims of the R01 research are beyond the scope of the R37 but will utilize longitudinal data collected as part of the R37 grant. No imaging data will be collected as part of this R01. Our first aim is to establish criteria for the definition of PiB+ and PiB- status in cognitively unimpaired elderly controls that are consistent with correlations of in vivo PiB retention and region-matched post- mortem correlates of A deposition determined for subjects who underwent PiB PET prior to death. The second aim will refine the PiB criteria and establish confidence intervals about the criteria that are consistent with longitudinal multi-modality neuroimaging results (i.e., presence/absence of AD-related imaging abnormalities). The final aim is to develop a working model of the natural history of in vivo amyloid deposition that accounts for spatial and temporal aspects of early amyloid deposition, using the results of Aims 1 and 2. The R37 imaging occurs at baseline and at follow-up intervals of 24 or 30 months, for subjects 65-84 years of age. The hypotheses will be addressed using statistical classification and modeling methods and neuropathological evaluations. The significance of the proposed research lies in the tools that will be developed to enable clinical researchers at our institution and throughout the world to improve the detection of earliest AD-related brain changes through better understanding of the implications of presymptomatic imaging abnormalities. Accurate definition of amyloid-bearing from amyloid-free individuals will be critical for the early identification of those who may benefit most from anti-amyloid therapy.